Quick and dirty script to copy files in AWS S3 to an S3 bucket in another AWS account with Node.js.

s3copy-across-aws-accounts.js

/*
 * This is a quick and dirty script for copying files in AWS S3 to other buckets.
 * This is written so it can work across AWS accounts.
 * You can hack in code to apply work or logic to each file.
 *
 * For each file, it downloads locally to a tmp file and then starts uploading as soon as possible
 * and deletes the file immediately when the upload is done.
 *
 * Best way to use: Spin up a micro in EC2, install node.js, 
 *                  create a directory and install the dependencies with 
 *                  npm install aws-sdk
 *                  npm install async
 *                  npm install node-uuid
 *                  then run with  node ./s3copy-across-aws-accounts.js.
 *
 * Dont Forget: 1. Make a folder named 'downloads' in the directory you will run from.
 *              2. Replace keys and bucket names in this file.
 *              3. Tweak any other code you want (simultaneous max uploads, downloads, etc)
 *
 * See also: AWS SDK:              http://docs.aws.amazon.com/AWSJavaScriptSDK/latest/frames.html
 *           Async for Node.js:    https://github.com/caolan/async
 *           Blog post about this: http://www.everyhaironyourhead.com/copy-files-in-aws-s3-to-an-s3-bucket-in-another-aws-account-with-node-js/
 */

var AWS = require('aws-sdk');
var async = require('async');
var fs = require('fs');
var uuid = require('node-uuid');

var sourceS3Conf = new AWS.Config({
  accessKeyId: 'SOURCEACCESSKEYGOESHERE',           //replace this with the S3 Access Key for the source bucket
  secretAccessKey: 'SOURCESECRETACCESSKEYGOESHERE'  //replace this with the S3 Secret Access Key for the source bucket
});
var sourceS3 = new AWS.S3(sourceS3Conf);

var destS3Conf = new AWS.Config({
  accessKeyId: 'DESTINATIONACCESSKEYGOESHERE',           //replace this with the S3 Access Key for the destination bucket
  secretAccessKey: 'DESTINATIONSECRETACCESSKEYGOESHERE'  //replace this with the S3 Secret Access Key for the destination bucket
});
var destS3 = new AWS.S3(destS3Conf);

var q = async.queue(upload, 5); //create an upload task queue that allows 5 simultaneous uploads

download(); //kick off the transfer

function download(marker){
  //listparams are the options for listing the source bucket contents. See AWS SDK.
  var listparams = {
    Bucket: 'SOURCE-BUCKET-NAME',  //replace this with the name of the source bucket
    EncodingType: 'url',
    MaxKeys: 100,      //Request 100 keys at a time
    Delimiter: '::'
  };
  //We request 100 bucket entry keys at a time, until we get through all of them.
  //AWS returns the starting key for the next 100 as NextMarker
  if(marker){
    listparams.Marker = marker;
  }
  sourceS3.listObjects(listparams, function(lerr, info) {
    if (lerr){
      console.log(lerr, lerr.stack);
      return; // if there is a problem getting the list, its all over.
    }
    //process the list of 100 keys (max) from the bucket
    async.eachLimit(info.Contents, 5, function iterator(item, cb){ // allow 5 simultaneous downloads
      //Add code here for processing the bucket entries. Tailor as needed.
      //For example, here we are skipping any entries that have a zero size.
      if(item.Size == 0){
        cb();
        return;
      }
      var getparams = {
        Bucket: 'SOURCE-BUCKET-NAME',  //replace this with the name of the source bucket
        Key: item.Key
      };
      sourceS3.getObject(getparams, function(gerr, s3obj) {  //download the object from the bucket
        if (gerr){
          //if there is problem just print to console and move on.
          console.log('Download issue with '+item.Key);
          console.log(gerr);
          cb();
          return;
        }
        var uname = uuid.v4(); //using uuid as a unique name for the tmp file we save
        //save the file to disk temporarily in a folder called downloads
        //until we upload it to the target bucket.
        //you can apply any rules or filters on the files here.
        fs.writeFile('./downloads/'+uname, s3obj.Body, {encoding: null}, function(fserr){
          if(fserr){
            //if there is problem just print to console and move on.
            cb(fserr);
            return;
          }
          //if everything went well downloading, add it to the upload queue.
          q.push({"filename": uname, "s3key": item.Key});
          console.log('Added '+item.Key+' to queue');
          cb();
        });
      });
    },
    function done(aerr){
      if(aerr){
        return console.log(aerr);
      }
      if(info.NextMarker){ //if there are more keys in the bucket, get the next 100
        process.nextTick(function(){
          download(info.NextMarker);
        });
      } else {
        console.log('Downloads Done! Please wait for uploads to finish.');
      }
    });
  });
}

function upload(uitem, qcb){ //function to upload each item in the upload queue to the destination bucket
  var stream = fs.createReadStream('./downloads/'+uitem.filename);
  destS3.putObject({
    Bucket: 'DESTINATION-BUCKET-NAME',  //replace this with the name of the destination bucket
    Key: uitem.s3key,
    Body: stream
  },
  function(uerr, data){
    if(uerr){
      //if there is an upload problem just print to console and move on.
      qcb(uerr);
      return;
    }
    //if the upload was successful, delete the local tmp file
    fs.unlink('./downloads/'+uitem.filename, function(delerr){
      if(delerr){
        //if there was a problem with that just print to console and move on.
        qcb(delerr);
        return;
      }
      console.log('Moved '+uitem.s3key);
      qcb();
    });
  });
}

Not able to download large files . Any suggestions?

Thank you very much! The script is solid. I added headObject to get the ContentType copied over as well.